Every estimating shop thinks of itself as selective. The numbers say otherwise. FMI's Contractor Business Performance Report has tracked bid-hit ratios for more than two decades, and the data pattern is stable: specialty subcontractors that average a 20-30% hit rate outperform shops at both extremes, and general contractors operating in hard-bid public work cluster tightly around a 10-15% hit rate on profitable work. A shop running at 45% is almost always leaving margin on the table by underbidding. A shop under 8% is burning estimating hours on the wrong pursuits.
The fix is not to bid more, and it is not to bid less. It is to bid the right ones. A weighted scorecard is the cheapest, least glamorous tool in a preconstruction department and the one that produces the biggest measurable improvement in annual gross profit. What follows is the seven-question rubric I have watched turn shops from 8% hit rate on break-even work to 24% hit rate on 6-point-margin work, in under a year.
The math that sets the target
Before the rubric, the math. The break-even bid-hit ratio is a function of your estimating cost, your expected margin, and your backlog constraint:
Break-even hit rate = (estimating cost per bid) / (expected gross profit per won job)
Plug in realistic numbers. Say a commercial electrical sub spends $4,800 in fully loaded estimating hours per medium-commercial bid (takeoff, coordination, RFIs, subcontractor quotes, assembly of the proposal). Say average gross profit per won job is $95,000. Break-even hit rate is 5.05%. Anything above that creates operating leverage. Anything below is a net cost with no revenue attached.
FMI's 2024 aggregated specialty-contractor data shows median bid volume of 62 pursuits per estimator per year and a median hit rate of 18.6%. That means the median estimator wins 11-12 jobs per year. A scorecard that lifts you from 18% to 24% is a 33% increase in annual revenue with the same estimating headcount. That is the prize.
The seven questions, in order of weight
Every question scores 0, 1, 2, or 3. Weights are fixed. The maximum raw score is 63. Pass/fail thresholds are calibrated to your backlog and cash position, but the defaults below are a reasonable starting point for most mid-market commercial shops.
| # | Question | Weight | Max |
|---|---|---|---|
| 1 | Owner/GC pay history (days-to-pay, disputes) | x3 | 9 |
| 2 | Schedule fit vs. current backlog | x3 | 9 |
| 3 | AR exposure if project goes sideways | x3 | 9 |
| 4 | Key personnel availability (PM + foreman) | x2 | 6 |
| 5 | Scope clarity and design maturity | x2 | 6 |
| 6 | Retention % and release triggers | x2 | 6 |
| 7 | Competitive field (number and identity of bidders) | x1 | 3 |
Decision bands: 45+ bid aggressively. 32-44 bid with a loaded markup. 20-31 bid only if strategic (key client, repeat work, learning curve). Below 20, no-bid and write the no-bid letter within 48 hours so you protect the relationship.
1. Owner/GC pay history
If the owner has a history of paying 72+ days or routinely withholds on small disputes, it is already losing money before the bid goes in. Score by actual days-to-pay on prior work: 0-45 days = 3, 46-60 = 2, 61-75 = 1, 76+ or any open dispute = 0. If you have never worked with the owner, pull a D&B report and a state lien search before you score. Default to 1 when data is missing.
2. Schedule fit vs. backlog
The schedule question is not "can we do it." It is "does this project's peak manpower requirement land on top of our existing peak?" If the two peaks overlap within 30 days, you are bidding a project that will force you to either decline work you already promised or hire into a peak that will disappear. Score 3 if peak demand lands in a valley, 0 if it stacks on the existing peak.
3. AR exposure
This is the one estimators skip. The AR exposure test: if the project stops paying tomorrow at the point of maximum cash-out (typically month 3 or month 4 of billing for a 9-12 month project), can your balance sheet absorb the loss? Run a 90-day cash-flow model on the project assuming net-45 pay terms and 10% retention. If the peak AR exceeds 20% of your available working capital, this is a 1. If it exceeds 40%, this is a 0 regardless of how good the job looks.
4. Key personnel availability
The named PM and named foreman, by start date. If either is still on a project that is behind schedule, you are bidding a job that will start with a B-team. Score 3 if both are clean, 1 if one is marginal, 0 if both are committed.
5. Scope clarity and design maturity
Use the AIA phase as a proxy. CDs at 95%+ with coordinated MEP backgrounds = 3. CDs at 75-90% = 2. DDs with performance specs = 1. SD or design-build with placeholder scope = 0. Design maturity correlates directly with change-order exposure.
6. Retention and release triggers
Contract retention at 5% with release at substantial completion is a 3. Retention at 10% with release at final acceptance (often 60-90 days after sub-complete) is a 1. Retention held through punch with a no-release-trigger clause is a 0. A 10% retention on a $2M job is $200k sitting with the owner for six months past your last labor spend.
7. Competitive field
Three to five qualified bidders = 3. Six to eight = 1. "Open to anyone" public bid with 12+ bidders = 0. The probability of winning on price alone drops geometrically with bidder count, and so does the probability that the winner has priced the job correctly.
Scoring the scorecard after the estimator has already started the takeoff. Sunk-cost bias will push every score up by one point. The rubric must be completed before a single hour is spent on the quantity survey, and a no-bid at that stage costs nothing.
A worked example
A 118,000 SF medical office fit-out. Owner is a regional hospital system we billed on two years ago; paid in 54 days, one dispute resolved in our favor. Schedule: peak labor lands in September, which is also when our federal courthouse job peaks. CDs at 85%. Named PM is the one we want, but he is still wrapping up a delayed project. Retention 10%, release at substantial completion. Six bidders listed on the plan-holder list.
| Question | Raw | Weighted |
|---|---|---|
| Pay history (54 days, clean) | 2 | 6 |
| Schedule fit (peak overlap) | 0 | 0 |
| AR exposure (22% of WC) | 1 | 3 |
| Key personnel (PM marginal) | 1 | 2 |
| Scope clarity (85% CDs) | 2 | 4 |
| Retention (10%, sub-complete) | 2 | 4 |
| Competitive field (6 bidders) | 1 | 1 |
| Total | 20 |
Score of 20 puts this on the no-bid line. The project is technically winnable, but the schedule fit and personnel constraints mean we would be bidding a job our organization can't execute without displacing higher-margin work. The honest call is to pass. Send the no-bid letter, stay on the bid list, and revisit the relationship on the next pursuit.
Calibrating the thresholds to your shop
The default 45/32/20 thresholds are a starting point. A shop with six months of backlog and limited PM bandwidth should tighten to 50/38/25. A shop coming off a revenue trough with estimators sitting idle can loosen to 40/28/17 for a quarter while it refills the pipeline. The important thing is not the specific thresholds but the discipline of applying them consistently across pursuits and tracking your hit rate by score band over time.
After 12-18 months of data, you will have an empirical hit-rate curve: percentage wins in the 45+ band, percentage in the 32-44 band, percentage below. The curve almost always slopes as expected, and once you have it, preconstruction turns into a managed portfolio instead of a guessing game.
"We cut our bid volume by 38% in one year and grew gross profit by 22%. The scorecard did both at once. Turning jobs down gets easier when the rubric has already told you why."
Operating VP, Top-50 ENR Specialty Contractor — Dallas, TX
Closing
The scorecard is not a replacement for judgment. It is a written record of judgment that your PMs, your CFO, and your next estimator can reproduce. The win-rate ceiling in construction is set by owner quality, schedule alignment, and cash discipline — not by takeoff speed. Every estimating shop that has quietly moved from 14% to 24% hit rate over the past five years has done it with some version of this rubric. Build yours.
PILRS automates the early-stage takeoff so your estimators spend their time on the scorecard, not on counting devices. See pricing here and stop bidding work you should have declined.